mtd/html index.html,1.29,1.30

Mon Sep 29 06:17:43 EDT 2003

Update of /home/cvs/mtd/html
In directory phoenix.infradead.org:/tmp/cvs-serv4577

Modified Files:
	index.html 
Log Message:
Update main web page. Remove ancient changelogs, update hardware support list slightly, mention IRC channel.


Index: index.html
===================================================================
RCS file: /home/cvs/mtd/html/index.html,v
retrieving revision 1.29
retrieving revision 1.30
diff -u -r1.29 -r1.30

--- index.html	8 Mar 2003 18:32:20 -0000	1.29
+++ index.html	29 Sep 2003 10:17:40 -0000	1.30
@@ -25,14 +25,16 @@
     </P>
 
     <HR size=2>
-    <H2>Mailing list</H2>
-    <P>There is a majordomo-managed mailing list: 
-      <B><I>lists-mtd at lists.infradead.org</i></B> - Full archives are available
-      at <A HREF="http://lists.infradead.org/pipermail/linux-mtd/">
-	http://lists.infradead.org/pipermail/linux-mtd/</A>. Now that I've
-      finally managed to get Mailman to display them properly, this includes
-      all the previous messages from before the changeover to Mailman.
+    <H2>Mailing list and IRC</H2>
+
 
+    <P>There is a mailing list for discussion of MTD development: <B><I><A
+    HREF="http://www.infradead.org/~dwmw2/email.html">linux-mtd at lists.infradead.org</A></i></B>. Please
+    do not post to the mailing list without first reading my <A
+    HREF="http://www.infradead.org/~dwmw2/email.html">notes on
+    etiquette</A>. Full archives are available
+      at <A HREF="http://lists.infradead.org/pipermail/linux-mtd/">
+	http://lists.infradead.org/pipermail/linux-mtd/</A>. 
       </P><P>
 
       To subscribe, go <A 
@@ -47,6 +49,9 @@
 	  HREF="mailto:linux-mtd-request at lists.infradead.org">
 	  linux-mtd-request at lists.infradead.org</A>
 	AS THE ABOVE SAYS.</B></P>
+
+    <P>There is also an IRC channel: <TT>#mtd</TT> on <TT>irc.ipv6.freenode.net</TT> or <tt>irc.freenode.net</tt></P>
+
     <HR size=2>
     <H2>Download and CVS</H2>
 
@@ -123,187 +128,6 @@
       used on it, along with a first-stage loader to load Grub itself into
       memory from the DiskOnChip, are in the CVS repository.</P>
 
-    <HR size=2>
-    <H2>TODO</H2>
-    <P>I'm impressed. The TODO list is getting much shorter.</P>
-    <uL>
-      <LI>Add an MTD driver for PCMCIA memory cards</LI>
-      <LI>Make the FFS2 code work on 2.3 kernels.</LI>
-      <LI>Implement XIP support in the 'map' layer and a read-only filesystem to use it.</LI>
-      <LI>Decrease the amount of flash space reserved by JFFS2 garbage collection
-    </uL>
-
-    <HR SIZE=2>
-    <h2>Changelog</H2>
-    <UL>  
-      <LI><B>Latest CVS</B><BR>
-	<UL>
-	  <LI>JFFS v2 -
-	    <UL>
-	      <LI>Compression</LI>
-	      <LI>Improved wear levelling - we deal with individual erase blocks now</LI>
-	      <LI>Improved concurrency - no longer all under one big semaphore</LI>
-	      <LI>Smaller RAM footprint - at least theoretically. And we forget stuff under memory pressure to be nice</LI>
-	      <LI>Hard link support</LI>
-	    </UL>
-	  </LI>
-	      
-	  <LI>New 'mtdblock' driver which allows you to use any normal
-	    filesystem on flash chips, albeit rather non-robust if you're
-	    writing to it. See <A HREF="#mtdblock">below</A> for more 
-	    information.</LI>
-	  <LI>New map drivers, including the Itsy (and iPAQ) driver.</LI>
-	  <LI>Massive NFTL overhaul. ECC, bad block handling, journalling fixes, etc.</LI>
-	  <LI>Verified DiskOnChip 2000 and Millennium support - some wonderful
-	    person seems to have gone through the code and the docs and ensured
-	    that we meet all the timing constraints that I ignored when I 
-	    threw the original code together.</LI>
-	  <LI>Support all sensible geometries for AMD/Fujitsu command set CFI flash</LI>
-	  <LI>Start (most?) of JFFS support for NAND flash</LI>
-	  <LI>Hardware support for a handful of NAND flash chips</LI>
-	  <LI>Beat JFFS up till it actually passes filesystem stress tests</LI>
-	  <LI>Allow CFI reads to suspend in-progress erases.</LI>
-	  <LI>Merged DiskOnChip 2000 and Millennium driver into one</LI>
-	  <LI>Partitioning of flash chips into multiple devices</LI>
-	  <LI>Map area drivers for various SBC boards</LI>
-	  <LI>New DiskOnChip Millennium driver, contributed by Ollie Lho</LI>
-	  <LI>Assorted JFFS fixes</LI>
-	  <LI>CFI code for AMD/Fujitsu Standard command set, mapped in 2x16 mode, contributed by Masami Komiya</LI>
-	</UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-20000704</B><BR>
-	Patch: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-20000704.diff">mtd-20000704.diff</A><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-20000704.tar.gz">mtd-20000704.tar.gz</A><BR>
-	<UL>
-	  <LI>Kernel patches for both 2.2.16 and 2.4.0 kernels to add MTD support, and
-	    a script file to move the source files into the right place.</LI>
-	  <LI>New <TT>map</TT> architecture for making mmap'd areas 
-	    available for use by generic chipset drivers</LI>
-	  <LI>Probe code for Common Flash Interface devices</LI>
-	  <LI>Run-time address alias detection for CFI chips</LI>
-	  <LI>Driver code for Intel Extended Command Set devices - connected
-	    any way you like as long as it's in 16-bit mode and not
-	    interleaved</LI>
-	  <LI>JFFS ported to 2.2 kernel</LI>
-	  <LI>Remove the /dev/mtd<I>x</I> character device driver from
-	    <TT>mtd.o</TT> and make it separately configurable as 
-	    <TT>mtdchar.o</TT></LI>
-	  <LI>Rename <TT>mtd.o</TT> to <TT>mtdcore.o</TT> just in case I
-	    didn't confuse enough people with the above</LI>
-	  <LI>Split the DiskOnChip probe routine into a separate file,
-	    allowing DiskOnChip 2000 and DiskOnChip Millennium drivers to
-	    be developed separately.</LI>
-	  <LI>Separate <TT>CONFIG_NFTL_RW</TT> option allowing you to
-	    compile safe read-only drivers (by turning it off)</LI>
-	  <LI>Imported Axis' <A 
-	      HREF="http://www.developer.axis.com/software/jffs/">Journalling
-	      Flash File System (JFFS)</A></LI>
-	  <LI>Updated FTL driver to the latest version from David Hinds' 
-	    <A HREF="http://pcmcia.sourceforge.org/">PCMCIA</A> code. Still
-	    some elusive bugs which cause it to corrupt the FTL - must get
-	    some PCMCIA flash hardware and check it worked <B>before</B> I
-	    started playing with it.</LI>
-	  <LI>Litter FTL driver with cpu_to_le16() et al calls in an 
-	    attempt to make it work on big-endian hardware</LI>
-	  <LI>Some cleanups to make the code conform to the <A HREF="tech/">new
-	      rules</A>.</LI>
-	  <LI>Add <TT>get_mtd_device()</TT> and <TT>put_mtd_device</TT> back
-	    again, having gone over the locking rules with a fine tooth
-	    comb and decided that they were OK if done properly.</LI>
-	  <LI>Added patch to GNU Grub which teaches it about the DiskOnChip.
-	    When it can also decode NFTL, it'll be more useful.</LI>
-	</UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-20000131</B><BR>
-	Patch: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-20000131.diff">mtd-20000131.diff</A><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-20000131.tar.gz">mtd-20000131.tar.gz</A><BR>
-	Complete patch against 2.3.41 kernel: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-20000131-against-2.3.41.diff.gz">
-	  mtd-20000131-against-2.3.41.diff.gz</A><BR>
-	<UL>
-	  <LI>Updated to work with 2.3.recent kernel. Apart from the obvious
-	    changes, this also includes turning off the block device request
-	    merging code, which seems to be duplicating requests rather than
-	    just merging them.</LI>
-	  <LI>Remove the byteorder workaround - there's a patch included which
-	    should fix the kernel.</LI>
-	  <LI>Added Grub loader for DiskOnChip - now someone just has to add
-	    DiskOnChip support to Grub itself.</LI>
-	</UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-20000119</B><BR>
-	Patch: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-20000119.diff">mtd-20000119.diff</A><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-20000119.tar.gz">mtd-20000119.tar.gz</A><BR>
-	<UL>
-	  <LI>NFTL now registers the block device <I>before</I> asking the 
-	    gendisk code to scan it for partitions - it works better like 
-	    that :)</LI>
-	  <LI>Various code cleanups, fixes for PPC and older kernels - mostly
-	  thanks to Pavel Roskin.</LI>
-	  <LI>No patch against 2.2 or 2.3.current kernels - I'll resume that 
-	    practice as soon as I've got more time to look after it.</LI>
-	</UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-19991025</B><BR>
-	Patch: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19991025.diff">mtd-19991025.diff</A><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19991025.tar.gz">mtd-19991025.tar.gz</A><BR>
-	Complete patch against 2.3.23 kernel: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19991025-against-2.3.23.diff.gz">
-	  mtd-19991025-against-2.3.23.diff.gz</A><BR>
-	<UL>
-	  <LI>Ported to 2.3 kernels (still compiles with 2.2)</LI>
-	  <LI>NFTL write support seems fairly reliable and interoperates with
-	    M-Systems' own drivers - more testing required</LI>
-	  <LI>Various bugfixes</LI>
-	</UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-19990820</B><BR>
-	Patch: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19990820.diff">mtd-19990820.diff</A><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19990820.tar.gz">mtd-19990820.tar.gz</A><BR>
-	<UL>
-	  <LI>Add <TT>nftl_format</TT> program 
-	    <I>(M-Systems' own drivers don't like the resulting NFTL yet)</I></LI>
-	  <LI>First version of write support for NFTL <I>(Likewise)</I></LI>
-	</UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-19990817</B><BR>
-	Patch: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19990817.diff">mtd-19990817.diff</A><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19990817.tar.gz">mtd-19990817.tar.gz</A><BR>
-	<UL>
-	  <LI>Support ECC for NFTL on DiskOnChip</LI>
-	  <LI>Behave correctly with multiple DiskOnChip</LI>
-	  <LI>Add primitive MAKEDEV script</LI>
-	  <LI>Move include directory to make it easier to compile into kernel</LI>
-	  <LI>Add <TT>/proc/mtd</TT> support</LI>
-      </UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-19990810</B><BR>
-	Patch: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19990810.diff">mtd-19990810.diff</A><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19990810.tar.gz">mtd-19990810.tar.gz</A><BR>
-	<UL>
-	  <LI>Various endianness and wordsize fixes</LI>
-	  <LI>Compile fixes for Linux/Alpha - DiskOnChip &amp; NFTL now works</LI>
-	  <LI>Sanity checks in NFTL driver to prevent oopsen on certain MTDs</LI>
-      </UL>
-	<BR>
-      </LI>
-      <LI><B>mtd-19990809</B><BR>
-	Tarball: <A HREF="ftp://ftp.linux-mtd.infradead.org/pub/mtd/mtd-19990809.tar.gz">mtd-19990809.tar.gz</A><BR>
-	<B>Large merge, including:</B>
-	<UL>
-	  <LI>Preliminary drivers for DiskOnChip 2000 and NFTL filesystem</LI>
-	  <LI>Microsoft FFS2 filesystem</LI>
-	  <LI>Octagon-5066 and vmax301 drivers for SBC boards</LI>
-	  <LI>Various other improvements from jgg.</LI>
-	  <LI>Boot loader code</LI>
-	</UL>
-      </LI>
-    </UL>
 
     <HR size=2>
 
@@ -342,18 +166,10 @@
 	working for both reading and writing, but could probably do with 
 	some more field testing before being used on production systems. 
       </LI>
-      <LI><B>Microsoft Flash Filing System</B><BR>
-	This provides a filesystem directly from the flash, rather than the 
-	less efficient filesystem-on-FTL-on-flash setup. The current driver is
-	in alpha-testing, but should allow read-only access to such 
-	filesystems.
-      </LI>
-      <LI><B>Journalling Flash File System</B><BR>
-	This was developed by <A HREF="http://www.axis.com/">Axis 
-	  Communications</A>, and provides a filesystem directly on the 
+      <LI><B>Journalling Flash File System, v2</B><BR>
+	This provides a filesystem directly on the 
 	flash, rather than emulating a block device. For more information,
-	see <A HREF="http://www.developer.axis.com/software/jffs/">Axis'
-	  developer pages</A>.
+	see <A HREF="http://sources.redhat.com/jffs2/">sources.redhat.com</A>.
       </LI>
     </UL>
     <HR SIZE=2>
@@ -367,47 +183,25 @@
 	memory with the linux-mtd system.
       </LI>
       <LI><B>PCMCIA devices</B><BR>
-	There are already a number of PCMCIA memory devices supported by Linux.
-	The MTD device subsystem is being designed with compatibility in mind,
-	so that the <A 
-	  HREF="http://hyper.stanford.edu/~dhinds/pcmcia/home.html">existing 
-	  PCMCIA drivers</A> can cleanly interface to the new MTD
-	subsystem. This work has not yet been undertaken, though.</LI>
+	PCMCIA flash (not CompactFlash but <em>real</em> flash) cards are
+	now supported by the <TT>pcmciamtd</TT> driver in CVS.
       <LI><B>Common Flash Interface (<A 
 	    HREF="http://www.pentium.com/design/flcomp/technote/cfi_1_1.htm">CFI</A>)
-	  onboard memory</B><BR>
-	This is one of the most interesting options for small embedded systems.
-	With a filesystem directly on the flash, rather than through an extra
-	layer like FTL, we can get execute-in-place (XIP) of programs and 
-	libraries on the flash, without needing to use up valuable system 
-	memory as a cache for the executable code.<BR>
-	A patch to the Linux kernel which supports this is available, and it 
-	should shortly be ported to the generic MTD system as a hardware
-	driver.</LI>
-      <LI><B>M-Systems' ISA Flash cards and DiskOnChip 1000</B><BR>
-	These devices are very simple - they seem to consist of a large array
-	of Intel flash memory, with a simple paging mechanism reminiscent of 
-	EMS, which allows different pages of the flash to be mapped into the 
-	device's 8K memory window. While working for Nortel a couple of years
-	ago, I developed drivers for these devices.<BR>
-	Unfortunately, I don't think the Disk-On-Chip 1000 is available any 
-	more.</LI>
+	  onboard NOR flash</B><BR>
+	This is a common solution and is well-tested and supported, most
+	often using JFFS2 or cramfs file systems.
+      </LI>
+      <LI><B>Onboard NAND flash</B><BR>
+	NAND flash is rapidly overtaking NOR flash due to its larger size 
+	and lower cost; JFFS2 support for NAND flash is approaching production
+	quality.
+      </LI>
       <LI><B>M-Systems' DiskOnChip 2000 and Millennium</B><BR>
-	M-Systems have now provided me with full documentation for the 
-	DiskOnChip 2000 and Millennium. The current driver is capable of
-	full operation on a DiskOnChip 2000, and I have received a
-	sample DiskOnChip Millennium from M-Systems, so I should get round to
-	adding Millennium support soon. The write support for NFTL appears
-	to pass my simple tests for interoperability - after I mount it with
-	my drivers, thrash it for a while, and unmount it, the NFTL is still
-	recognised by M-Systems' own drivers. More testing of this would be
-	very much appreciated.
-	<BR>
-	<I>The <A HREF="doc2000.html">Linux and Disk-On-Chip
-	  2000 page</A> is available here. Please note that, now that 
-	M-Systems have provided me with complete specifications with which to
-	write a free driver for the device, there is no longer any reason 
-	to avoid using it with Linux systems.</I>
+	The DiskOnChip 2000, Millennium and Millennium Plus devices should
+	be fully supported, using their native NFTL and INFTL 'translation
+	layers'. Support for JFFS2 on DiskOnChip 2000 and Millennium is 
+	also operational although lacking proper support for bad block 
+	handling.
 	<HR>
 	</LI>
 


